Boillot, G., Winterer, E. L., et al., 1988 Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 103

11. LATE JURASSIC CARBONATE PLATFORM OF THE GALICIA MARGIN1

L. F. Jansa, Atlantic Geoscience Centre, Geological Survey of Canada, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, Canada M. C. Comas, Instituto Andaluz de Geologia Mediterranea, C.S.I.C., Universidad de Granada, Campus Universitario de Fuentenueva, Granada, Spain M. Sarti, Dipartimento di Scienze della Terra, Universita della Calabria, Cosenza, Italy and J. A. Haggerty, Department of Geosciences, The University of Tulsa, Tulsa, Oklahoma

ABSTRACT The carbonate platform of Tithonian-?Berriasian age encountered on the western flank of the Galicia Bank at Site 639 is part of an extensive Late Jurassic carbonate shelf extending from Algarve to the Lusitanian Basin, Galicia Bank, and eastern Grand Banks of Newfoundland. The platform, which locally could be up to 400 m thick, probably overlies older sedimentary or low-grade metamorphic strata at Site 639. The 287 m of carbonates penetrated at this site was de• posited mainly in shelf lagoon and shelf edge shoals, with localized low-relief biohermal mounds, in response to eu• static sea-level fluctuations and regional subsidence during the Tithonian-Berriasian. Minor block faulting influenced lithofacies distribution at the shelf edge. Shallow burial cementation was the most significant diagenetic event. After burial by several hundred meters of younger sediments, the top of the carbonate platform underwent extensive dolomiti- zation during a period of tectonic fragmentation, followed by subaerial exposure and erosion of the platform. Such pro• cesses resulted in a complex carbonate diagenetic history with several periods of dolomitization. The major dolomitiza- tion event is the result of the replacement or mixing of formation fluids with evaporitic brines in a shallow subsurface setting. The emerged carbonate platform drowned during the middle early Valanginian as a result of eustatic sea-level rise and tectonic subsidence and was covered by Valanginian turbidites. During a second, more intensive period of tec• tonic deformation, probably in the late Valanginian, the development of large rotational faults resulted in tilting of the carbonate platform blocks toward the east. After tilting, the outer margin rapidly subsided to abyssal depths during the Late Cretaceous-early Tertiary. The development of the carbonate platform helps in dating the major tensional episodes on the western flank of Galicia Bank as late Berriasian-earliest Valanginian and ?late Valanginian in age.

INTRODUCTION The finding of Lower Cretaceous terrigenous turbidite beds beneath seismic reflector 4 (Fig. 1C), previously identified as the The depositional and tectonic history of the Late Jurassic top of the pre-rift carbonate platform sequence at Site 638 carbonate platform on the western side of Galicia Bank is pre• (Mougenot et al., 1985; Boillot, Winterer, et al., 1987), required sented in this contribution. The results of our study have pro• major revision of the existing seismic interpretation. Drilling vided new evidence about the tectonic evolution of the Galicia difficulties in the Lower Cretaceous turbidite sequence at this margin and have advanced our understanding of the paleogeog• site led to the abandonment of the hole without reaching the raphy of the northern central North Atlantic during the Late Ju• deep objective. rassic-Early Cretaceous. Shipboard reassessment of the seismic data showed that the This chapter is divided into two parts. The first, analytical carbonate platform could be reached about 3 km to the west of part provides a detailed petrographic description of the sedi• Site 638, where it possibly could be exposed at the fault escarp• mentary rocks encountered. The second part is data interpreta• ment, covered by a thin veneer of Cenozoic sediments. Thus, tion, a discussion of paleoenvironment, diagenesis, dolomitiza• Site 639 was established to drill through the carbonate platform tion, tectonics, and paleogeography. into underlying basement rocks (Fig. 1C). Technical difficulties The objective of Ocean Drilling Program (ODP) Leg 103 was in drilling the carbonates forced premature abandonment of to resolve the geologic and paleoceanographic history of the Hole 639A after penetrating only 89.8 m below seafloor (mbsf)- Galicia margin (Fig. 1A) in relation to the rifting and continen• In an effort to reconstruct the stratigraphy of the carbonate tal separation between the Iberian and eastern North American platform, another five holes were drilled in an east-west transect continental plates. A multiple reentry hole was located on re• about 800 m long, downdip of the escarpment (Fig. 2) (Ship• gional seismic line GP-101 (Fig. IB) and designated as ODP Site board Scientific Party, 1987b). The reason for this six-hole tran• 638 (Fig. 1C). The site was located near the upper part of the sect was that none of the holes was able to penetrate into the upward tilted edge of the rotated tectonic block to ensure sam• carbonates by more than few cores before the drill pipe would pling of a complete Mesozoic stratigraphic succession from rift• become stuck, forcing abandonment of the hole. ing phases, through the drowning of the carbonate platform, to Drilling operations at Site 639 provided only fragmentary data changes in the environment during plate separation as the mar• about the carbonate platform and its history (Shipboard Scien• gin thermally subsided to oceanic depths. tific Party, 1987b). The incomplete stratigraphic record together with poor core recovery (about 20% in Holes 639A through 639D and less than 5% in Holes 639E and 639F) resulted in an uneven 1 Boillot, G., Winterer, E. L.( et al., 1988. Proc. ODP, Sci. Results, 103: Col•sample coverage for reconstruction of the carbonate platform lege Station, TX (Ocean Drilling Program). depositional history.

171 L. F. JANSA ET AL.

Figure 1. A. Location of Site 639 on the Galicia Bank. B. Sea Beam map of the area with locations of Site 639, seismic sections, and dredge stations. Note that the tectonic structure is mimicked in the seafloor morphology with the upward-tilted edges of the half-grabens forming steep escarpments at the sea bottom. The two dredge stations on these escarpments are DROl and DR03. Approximate locations of submarine dives (Boillot et al., this volume) are marked by circled numbers. C. Multichannel seismic-reflection line GP-101 with plotted location of Site 639. The numbers on the line correspond to the seismic data interpretations prior to drilling (Mougenot et al., 1985; Boillot, Winterer, et al., 1987). According to this interpreta• tion, seismic Unit 5 corresponds to the pre-rift sequence, Unit 4 to syn-rift, and Unit 3 to post-rift strata. It was assumed that the top of the carbonate platform corresponded to seismic reflector 4B. Frame outlines portion of seismic line geologically interpreted in Figure 2.

The Shipboard Scientific Party (1987b) proposed two possi• ume) complements our study. However, there is major differ• ble interpretations for the seismic section across Site 639. In this ence between our and Moullade et al.'s studies in that the latter paper we support and modify the normal-fault interpretation authors used as a working hypothesis the early concept of the (Fig. 47 of the Shipboard Scientific Party, 1987b) for the sea• shipboard party (Fig. 27 of Shipboard Scientific Party, 1987b), ward edge of the carbonate platform. That interpretation im• according to which the holes tested progressively deeper strati• plies that drilling a transect across the buried carbonate escarp• graphic horizons of the carbonate platform down to the under• ment may not provide a complete stratigraphic section through lying basement. This difference in concept of tectonic setting re• the platform (Fig. 2). sults in different interpretations not only of the carbonates but Considering the poor core recovery, the interpretation of the also of the tectonic evolution of the whole region. Mesozoic carbonate platform that we provide in this paper is generalized and subjective. It is based on core and thin-section TECTONIC SETTING studies, geophysical logs, and seismic-reflection data. The known Detailed discussion of the tectonic setting of the Galicia evolution of synchronous carbonate platforms in northern Spain, Bank can be found in the Initial Reports of Leg 103 (Boillot, Portugal, and the Grand Banks and the results of dredging on Winterer, et al., 1987). Here, we concentrate only on the part the western Galicia margin (Dupeuble et al., 1987) assisted in re• relevant to the development of the carbonate platform. Site 639 constructing the geologic history of the Galicia Bank carbonate is on the uplifted, outer flank of a tilted block (Fig. 1C). The platform. section dips to the east and subcrops on a westward-sloping sub• We dwell only briefly on limestone diagenesis and dolomiti- marine erosion surface, now buried under an onlapping wedge zation, the subjects of other contributions in this volume (Lo• of Upper Cretaceous and Cenozoic pelagic sediments. The top reau and Cros; Haggerty and Smith; Daniel and Haggerty). The of the carbonate platform is overlain unconformably by Valan• biofacies study of the limestones by Moullade et al. (this vol• ginian pelagic sediments.

172 LATE JURASSIC CARBONATE PLATFORM

639 FE D C BA 638

10 km

^-rrTJv^l.^-^ VV' Basement >>''- ' ' ' 8J Figure 2. Geologic interpretation of the segment of seismic line GP-101 shown in Figure 1C, demonstrating the tectonic setting of the carbonate plat• form near Site 639 and position of drill holes. Seismic reflector C corresponds to the top of carbonate platform; reflector V to an untested clastic unit of probable Hauterivian age.

Reinterpretation of the seismic-reflection profile across Site ume). The strong reflector that occurs at 8 s two-way traveltime 639 (Fig. 2) shows that the outer edge of the carbonate platform (twt) below seafloor on seismic line GP-101 near the drill site is highly broken by synthetic and antithetic faults. Faulting re• (see Mauffret and Montadert, this volume) we interpret as cor• sulted in downthrown blocks to the west, with rotation along lis- responding to the top of the crystalline basement (Fig. 2). tric faults. Small tectonic troughs produced between the fault Combining the results of seismic data interpretation, drill• blocks acted as catchment basins for the upslope-derived detri• ing, and studies of thin sections, we have concluded that Hole tus. 639A was drilled at the top of the block-faulted carbonate plat• The carbonate platform seems to thicken to the west, with form (Fig. 2). Hole 639B was probably drilled into the same tec• the thickest section near the escarpment edge where the seismic tonic block but encountered a deeper stratigraphic level than thickness reaches approximately 0.17 s (Fig. 1). Applying an av• Hole 639A. Hole 639C is situated in one of the tectonic grabens erage compressional seismic velocity for carbonates of 4.7 km/s between rotated limestone blocks and thus, recovered debris de• (Shipboard Scientific Party, 1987b), the thickness of platform rived probably from the upslope exposure of basement. We sug• could be up to 420 m. However, because the platform is inho- gest a similar origin for the polymictic debris of elastics, car• mogeneous from intercalation of elastics, the thickness is possi• bonates, and basement rocks encountered in Holes 639E and bly less. The strong seismic reflector at the base of the outer• 639F, which are in a similar tectonic depression but below the most downfaulted block points to separation of the carbonate next downslope rotated limestone block. Hole 639D was proba• platform from the underlying rocks by an unconformity. The bly drilled into one of these limestone blocks and provides the seismic data do not adequately resolve the character of the rocks only stratigraphically continuous section of the carbonate plat• underlying the carbonate platform. The character of the seismic form. However, the carbonate platform was not penetrated in its sequence below the carbonate platform could indicate that it is entirety at this site. either of sedimentary origin or is composed of low-grade meta• sedimentary rocks. The thickness of this basal sequence, under• CARBONATE PLATFORM lain by a weak reflector, is approximately 1400 m (assumed com• Holes 639D, 639B, and 639A have been used to reconstruct a pressional velocity of 4 km/s). Such an interpretation seems to composite stratigraphic section of the carbonate platform (Fig. find support in the results of submersible dives along the Gali• 3). The limestones in Holes 63 9A and 63 9B are completely dolo- cia margin (Boillot et al., this volume). Dives on the escarpment mitized; in Hole 639D only the top is dolomite. For descriptive 6 km to the north of the drill site documented the presence of purposes, we retained the general lithologic framework subdivi• about 500 m of sedimentary rocks that are older than those re• sion established by shipboard study (Shipboard Scientific Party, covered at Site 639. This sequence is composed of intercalated 1987b). The Valanginian marlstones overlying the carbonate plat• elastics and dolomites, with dolomites near the base. The age is form at Hole 639A correspond to Unit III, the underlying dolo• unknown but is possibly Mesozoic. The sediment sequence over• mites to Unit IV, and the limestones of Hole 639D to Unit V. lies weakly metamorphosed sedimentary rocks intercalated with The units are described in ascending stratigraphic order in the volcaniclastics of probable Paleozoic age (Boillot et al., this vol• following text. The petrographic data for the limestones and do-

173 L. F. JANSA ET AL.

SITE 638 LEGEND Hole 638 B and C (Sea floor -4673m) ^ 18

Hole

100 200 300 Metres Figure 3. Composite stratigraphic section of Sites 639 and 638, showing schematic lithology and lithostratigraphic subdivi• sion of drilled sedimentary strata. Water depths of each hole are in meters below sea level, as determined from echo-sound• ing profiles.

174 LATE JURASSIC CARBONATE PLATFORM lomites are summarized in geologic charts (Fig. 13), which are ules, gastropods (Nerinea), ostracodes, and mollusk debris. Some similar to those published for Jurassic limestones off Morocco dasycladacean algae debris is also observed (Fig. 4C). Traces of by Jansa et al. (1984). calcisponges, corals, and hydrozoans are also found in this sub• unit. The allochems are enveloped by micrite. From the lower Unit V part of Core 103-639D-11R to the base of the drilled sequence, Unit V, composed of limestones and intercalated elastics, the micritic matrix shows crystal enlargement (Fig. 4D), in which was encountered in Hole 639D. The thickness of the penetrated the micrite is replaced by microspar. The molds from dissolved sequence is 94.1 m. Three subunits are recognized on the basis skeletal fragments and voids in some of the skeletal debris are of changes in lithological composition (Shipboard Scientific filled by slightly ferroan, coarse, blocky sparry calcite. Dolo• Party, 1987b): the basal Subunit VC is characterized by the oc• mite occurs only in traces along some of the stylolitic seams and currence of oncolith-skeletal wackestones and packstones; Sub• fracture fillings and also as reliefs in the matrix. Ghosts of dolo• unit VB consists of mixed terrigenous sandstone-marlstone and mite rhombohedra suggest that minor dedolomitization oc• fine-grained limestone lithologies; and the upper Subunit VA is curred. biohermal limestone beds intercalated with marlstone and skele- Skeletal-Pelletoidal Packstone Microfacies. This microfacies tal-peloid wackestone containing planktonic biota. Each of the is found mainly in the upper part of Subunit VC (Cores 103- subunits includes a variety of microfacies, which we describe 639D-9R to 103-639D-11R; Fig. 3), where it is intercalated with here in more detail. the oncolith-skeletal-peloid packstone microfacies. The only dif• ference between these two microfacies is that this one lacks on• Subunit VC: Large Foraminifer-Oncolith Limestone coliths. Small foraminifers (valvulinids and Trocholina) and Lithofacies (base of Core 103-639D-13R to top of Core mollusk and dasycladacean algae debris are the recognizable bi• 103-693D-9R) ota in the packstones. In some of the packstones, the peloids The limestones in Subunit VC are light to brownish gray on- were merged by compaction, which gives the rocks an appear• colithic wackestones and packstones and peloidal packstones, ance of being dominantly micrite. However, closer examination intercalated with a minor amount of gray, skeletal peletoid wacke• reveals the original peloid nature of the sediment. The merged stone to grainstone. Two thin beds of pale olive marlstone are peloid texture indicates that these peloids were of a soft type intercalated with the limestone in Core 103-639D-10R. The marl• and that lithification had to occur relatively early. The peloids stone is bioturbated and contains scattered foraminifers and are of a different origin than the "hard" peloids of the oncolith rare serpulids. Limestones are fractured with hairline fractures microfacies. filled by calcite. Stylolites are commonly oriented parallel to Fossiliferous Micrite Microfacies. Except in Section 103-639D- bedding planes; locally, development of horsetail stylolites re• 10R-1, this microfacies is otherwise only a minor component of sults in a pseudobrecciated limestone texture. The intensity of Subunit VC. Silt- to fine-sand-size bioclast debris represented by pressure solution is demonstrated by the juxtaposition of very large foraminifers, thick-walled mollusk shells, gastropod and different lithologies on each side of the stylolite, as seen in Sec• dasycladacean algae debris, and serpulids (Figs. 4E and 4F) is tion 103-639D-13R-2. Concentration of iron oxides along some surrounded by a matrix of micrite, with traces of quartz silt and of the stylolite seams suggests that the stylolitic seams were pas• clay. The generally finer-grained texture of the limestones in sageways for the migration of oxygen-rich fluids. Cores 103-639D-10R and 103-639D-9R and the appearance of marls in Core 103-639D-10R are indicative of a transitional Microfacies Description boundary between Subunits VC and VB and of a decrease in en• Oncolith-Skeletal-Peloid Wackestone to Packstone Microfa• ergy conditions. cies. Packstone is the most common textural type in Subunit VC and consists of a mixture of oncoliths, skeletal debris, peloids, Subunit VB: Mixed Terrigenous Clastics-Limestone and micrite. The most characteristic feature of these rocks is Lithofacies (Cores 103-639D-8R through 103-639D-6R) their grain size bimodality (Fig. 4A). The coarse fraction is com• Light gray, light yellowish brown to light reddish brown fine• posed of oncoliths 0.1 to 4 mm in diameter, averaging about grained limestones are intercalated with about 20-cm-thick beds 2 mm in diameter. The second, finer fraction is of fine-sand-size of light yellowish brown silty marlstone, silty calcareous clay, peloids. Oncoliths (l%-40%) are scattered throughout the sedi• and olive-colored clay. Intercalations of a coarse-grained, yel• ment and are only rarely concentrated in thin laminae. The on• lowish brown sandstone and pale olive siltstone were sampled in coliths have nuclei composed of skeletal particles (larger fora• Cores 103-639D-7R and 103-639D-8R (Fig. 5). The contacts be• minifers, algal debris, and mollusk debris), composite grains, tween individual lithologies were generally not recovered, with micrite, or sparry calcite, which fills a cast in replacement of the exception of some gradational contacts with marls in Cores dissolved skeletal debris. Sessile foraminifers locally overgrow 103-639D-6R and 103-639D-8R. Density and gamma-ray logs the oncoliths. The finely laminated texture of the original algal indicate that the sandstone beds are more common in this unit coatings is rarely preserved. Almost all of the oncoliths show a than is apparent from the coring results. The terrigenous beds high degree of micritization, which mostly affected the outer on the geophysical logs are 1 to 2 m thick and alternate with 1- envelopes, thereby giving the appearance of a massive micrite to 9-m-thick limestones. The limestones are of variable compo• rim (Fig. 4B). sition, with the most common type being sandy quartz or silty Peloids are the other dominant component of this microfa• quartz skeletal-peloid packstone to wackestone (Fig. 6A). In mi• cies, accounting for up to 50% of the sediment. The peloids are nor abundance are skeletal wackestone, biomicrite, and skeletal- 0.07 to 0.25 mm in diameter, rounded to oval in shape, with a oncolith wackestone, along with rare occurrences of foraminif• well-defined boundary (Fig. 4A), and were of a hard type. Ghosts eral grainstone. of a skeletal fabric are observed in the center of some larger pe• Macroscopic examination shows locally intensive stylolitiza- loids, which suggests that some of them are micritized skeletal tion of the limestones. Bedding is not well developed, but in debris. Others, particularly the oval-shaped larger peloids, could some of the marlstones, lamination is caused by the accumula• be fecal in origin. Fossil remains (up to 10%) are dominated by tion of large foraminifers into 5- to 20-mm-thick laminae (Fig. large benthic foraminifers, Anchispirocyclina lusitanica and Litu- 5). These laminae have a 20° to 30° inclination that may be in• olidae (Moullade et al., this volume); less common are small dicative of the magnitude of post-depositional tilting of the benthic foraminifers {Trocholina and Lenticulina), echinoid spic• beds.

175 L. F. JANSA ET AL.

Figure 4. Thin-section photomicrographs of Subunit VA. A. Oncolith-peloid wackestone to packstone. Note the bimodality in size distribu• tion of the particles. The coarse fraction consists of oncoliths that have nuclei of large foraminifers and mollusk and unidentifiable skeletal debris. The finer fraction is comprised of mostly subangular, less rounded, hard peloids. Sand-size bioclast debris and small benthic fora• minifers (valvulinids) are less common. Crystal enlargement resulted in replacement of the original micrite matrix by microspar, which pro• duced a pseudograinstone texture. Sample 103-639D-13R-1, 29 cm; ordinary light; scale bar = 1 mm. B. Detail of an oncolith showing the intensively micritized outer rim. Inside the oncolith the laminar character of the coatings around the gastropod nucleus is still recognizable. Sample 103-639D-13R-2, 72 cm; ordinary light; scale bar = 1 mm. C. Dasycladacean algae (Linoporella and Epimastopora) debris in an on• colith-peloid packstone. Sample 103-639D-12R-3, 56 cm; normal light; scale bar = 0.2 mm. D. Well-sorted skeletal-peloid packstone. Small foraminifers dominated by valvulinids are common; other skeletal debris is poorly preserved and micritized. Hard peloids are mostly suban• gular in shape. Micritic matrix has been replaced by microspar. Sample 103-639D-11R-2, 45 cm; ordinary light; scale bar = 0.5 mm. E. Large foraminifer encrusted by serpulids in biomicrite. Fragments of algae occur at lower left. Sample 639D-10R-1, 74 cm; crossed nicols, scale bar = 1 mm. F. Biomicrite cut by two diachronous sets of sparry calcite-healed fractures that intersect at a 45° angle. Sample 103- 693D-10R-1, 128 cm; normal light; scale bar = 1 mm.

176 LATE JURASSIC CARBONATE PLATFORM

cm

Figure 5. Typical lithology of Subunit VB. Clayey limestones alternate with beds of marlstone and me• dium- to coarse-grained sandstone in the 40-80 cm interval of Sections 103-639D-7R-1, 103-639D-7R-2, 103-639D-7R, CC, and 103-639D-8R-1.

177 L. F. JANSA ET AL.

Microfacies Description micritized oolitic grains cemented by blocky, low-iron sparry Sandy Quartz and Silty Quartz Limestone Microfacies. A calcite. distinct feature of this microfacies, other than the presence of Subunit VA: Biohermal Limestone Lithofacies (Cores terrigenous components, is the occurrence of a broad variety of 103-639D-5R through 103-639D-4R) small benthic foraminifers. Particularly common are epistominids and litulidae; lagenids, miliolids, and valvulinids are rare. A. lu- Geophysical logs indicate that this subunit consists of about sitanica is present in the skeletal-oncolith wackestone microfa• 4 m of well-lithified limestone, which was penetrated by Core cies, which is similar to that in Subunit VC, and was found to be 103-639D-5R, overlain by a marlstone with another 1-m-thick present in all Subunit VB cores. Of the other biota, gastropods limestone found in Core 103-63 9D-4R. However, only 3% of and thick-walled mollusks are common. Algae debris is a minor Core 103-639D-4R was recovered, consisting of pieces no larger constituent (Fig. 6B), dominated by dasycladacean algae (up to than 6 cm of limestone of varying colors and lithologies. This 5% in some cores), with Cylindroporella, Salpingoporella annu- may indicate that the limestone in Core 103-639D-4R is a brec• lata, and Zergatella identified. Also occurring are Cyanophytae cia from the exposed limestone escarpment above, but it may (Girvanella type and spongiostromata), which coat bioclasts also reflect the lithologic variation of the strata. No soft sedi• and construct incipient stromatolite-like growth structures (Fig. ment, which may have separated limestone beds, has been re• 6C). Annelid, echinoid, and ostracode debris is rare, and coral, covered; thus, it remains unknown whether the limestone in calcisphere, and calpionellid debris is found in trace amounts. Core 103-639D-4R is an autochthonous or allochthonous debris Algal oncoliths about 3 mm in diameter are a minor component deposit. The similarity in composition between the limestones of a skeletal-peloid limestone in Cores 103-639D-6R and 103- in Cores 103-639D-4R and 103-639D-5R and regional consider• 639D-8R. Oncoliths in Core 103-639D-8R have an irregular out• ations lead us to consider the limestones to be autochthonous line coated by iron-manganese film, which is indicative of a deposits. Because only fragments of variable composition were brief period of nondeposition and solution, most probably as a recovered, we have not divided this subunit into microfacies. result of a periodic influx of fresh water. Strong micritization of The limestone in Core 103-639D-5R consists of fragments of bioclasts in the mixed carbonate/terrigenous elastics microfa• a hydrozoan floatstone, framestone, and rudstone (Figs. 7A cies also suggests alteration resulting from fresh water. Silt-size through 7C) intercalated with intraclast skeletal packstone, skel• peloids and micrite (which has been locally recrystallized into etal wackestone, packstone, and biomicrite. The biohermal lime• microspar) plus minor clays and quartz silt to fine-grained sand stone is fractured (Fig. 7C), with fractures filled by iron oxide- form the matrix of the packstones and wackestones. stained sparry calcite. The fauna of the biohermal limestone is a mixture of hydrozoans (Parastroporoidea), chaetetids (Blasto- Foraminiferal Grainstone Microfacies. This microfacies was chatestes; Figs. 7A and 7B), stromatoporoids (Cladocoropsis recovered in Sample 103-639D-7R-1, 71-75 cm. The grainstone miriabilis), corals (Stylamina and Pennulados), bryozoans, and is well sorted and reversely graded. The most common grains calcareous sponges. Several stages of encrustation by Girvanella, are rounded or lightly coated debris of large foraminifers (15%- Thaumatoporella, Lithocodium, and microbial crusts can be 20% A. lusitanica), worn mollusk and gastropod debris (some recognized on surfaces of some of the hydrozoans (Fig. 7A), pieces of which were micritized and encrusted by sessile foramini• suggesting multiple encrustations and organic binding of frame- fers), poorly preserved red algae (Lithothamnium), and dasycla• building biota. The debris of a frame-building biota were depos• dacean algae (Cylindroporella). Minor components are fine quartz ited in bioturbated, silt-size fossiliferous-peloid carbonate muds, grains and rounded micrite intraclasts, all cemented by patchy, with echinoids, mollusk debris, and benthic foraminifers (Fig. sparry-microsparite calcite (Fig. 6D). Grainstone is a rare micro• 7B) as additional skeletal carbonate constituents. facies not only in Subunit VB but in all of Unit V. The biohermal limestone is intercalated with a skeletal-pe• Terrigenous Sediments. All types of gradations between clean loid wackestone and packstone similar in composition to that carbonate, marl, and calcareous to terrigenous sand are found forming the "matrix" of the biohermal limestone. Preservation in Subunit VB. In the marl, clay input dilutes the carbonate of geopetally filled borings in the skeletal-peloid wackestone without observably changing its texture. Concentrations of larger (Sample 103-639D-5R-3, 42 cm) requires early lithification. A foraminifers into laminae in the marls indicate redeposition, cryptalgal, clotted structure resembling a thrombolite in a frag• current transport, and mixing of sediment from two sources. mented skeletal-peloid packstone was observed in Sample 103- Terrigenous sandstones are fine- to medium-grained and mostly 639D-5R-2, 95 cm (Fig. 7E). Such features are occasionally poorly sorted, with grains angular to subangular in shape (Fig. found to grow between skeletal debris in the reef flank deposits 6E). The dominant constituent is quartz, with feldspars as a mi• in a deep sub tidal environment. The accumulation of algal de• nor component (6% to 16%). The feldspars range from fresh to bris, belonging to Solenoporacea [Thaumatoporella), Gymno- almost completely kaolinized. Plagioclase is the most common, codiacea (Permocalculus), and Coadiaceae (Arabicodium, with microcline and orthoclase less common. Chert, sericitic Lithocodium agregatus, and Bacinella irregularis) (Sample 103- quarzite, metamorphic quartz, chloritized biotite, and musco• 639D-5R-3, 34-43 cm) contrarily indicates a shallower deposi• vite are the other minor constituents of the sandstone. Felsic ig• tional environment, unless the algal debris was retransported. neous rock fragments occur in trace amounts. Of special signifi• As a diagenetic curiosity, we note the occurrence of corals, re• cance for elastics provenance determination is the clast of feld• placed in part by coarse, zoned dolospar and silica (Sample 103- spathic rock in the sandstone of Section 103-639D-7R-2. The 639D-5R-2, 75-77 cm). feldspars in this lithoclast have a radial fibrous texture, similar As no samples were available for study of the 10-cm-thick to that found in "basement" rock fragments in Section 103- bed of olive silty clay intercalated with the limestones at 40-50 639F-2R, CC. Some of the feldspathic sandstones (Section 103- cm in Section 103-639D-5R-1, the composition and origin of 639E-7R-2; Fig. 6E) lack carbonate grains, have an argillaceous this bed is unknown. matrix, and are cemented in part by coarse dolospar. Other feld• At the top of Subunit VA, the limestone is pale, yellowish spathic sandstones (Section 103-639E-8R-2) are better sorted and brown and pink, skeletal peloid wackestone and biomicrite. The contain up to 10% carbonate allochems composed of rounded individual core pieces recovered are not larger than 6 cm long sand-size clasts of micrite, rare worn mollusk debris, larger for• and are rather rounded. Fauna recognizable in this limestone in• aminifer tests (Fig. 6F), micrite-coated quartz grains, and a few cludes unsorted sand-sized debris of mollusks, gastropods, bryo-

178 LATE JURASSIC CARBONATE PLATFORM

Figure 6. Thin-section photomicrographs of Subunit VB. A. Quartz silty skeletal wackestone. Holothurian plate in center. Mollusk debris and epistominids are coated by thin micritic envelopes and surrounded by a peloid, quartz silt, micrite matrix. Sample 103-639D-8R-1, 25 cm; normal light; scale bar = 1 mm. B. Unsorted skeletal packstone with common, small benthic foraminifers (Epistomina) and dasycla- dacean algal debris. This sample also contains a preserved fragment of Permocalculus. Oncolith with a mollusk fragment nucleus occurs at the top of view. Sample 103-639D-6R-1, 60 cm; ordinary light; scale bar = 0.5 mm. C. An algal stromatolite began to develop as an overgrowth on a gastropod fragment. The rest of the limestone is a skeletal wackestone with merged silt-size peloids and an argillaceous micrite matrix. Sample 103-639D-6R-1, 90 cm; ordinary light; scale bar = 1 mm. D. Foraminiferal grainstone. The dominant skeletal components are large foraminifers (Anchispirocyclina lusitanica) and a minor amount of abraded, thick-walled mollusk shell fragments cemented by sparry calcite. Note the reverse grading in the sediment. Sample 103-639D-7R-1, 71 cm; ordinary light; scale bar = 1 mm. E. Unsorted, medium-grained feldspathic sandstone with grains mostly subangular to angular in shape. Feldspars show a variable degree of alteration. The large orthoclase grain in the center is sheared and slightly kaolinized. The sandstone matrix is argillaceous with dolospar cement. Sample 103-639D-7R-2, 53 cm; polarized light; scale bar = 0.5 mm. F. Large foraminifer test and micrite intraclasts in a fine• grained calcareous sandstone. Sample 103-639D-7R-1, 34 cm; normal light; scale bar = 0.5 mm.

179 L. F. JANSA ET AL.

Figure 7. Thin-section photomicrographs of Subunit VA. A. Framestone formed by a complex overgrowth of hydrozoans and algae. A coral fragment is overgrown by the alga Bacinella irregularis, which is in turn coated by a rim of Thaumatoporellaporovesiculifera. This latter alga is overgrown by chaetetids (Shugraia) encrusted by Lithocodium, which formed a substratum for a new colonization by chaete- tids. Sample 103-639D-5R-3, 42 cm; ordinary light; scale bar = 0.5 mm. B. Breccia of biohermal limestone, composed of clasts of chae• tetids, thick-walled mollusk shell fragments, echinoids, and calcisponges (the latter component not photographed). Fractures are healed by coarse, blocky sparry calcite. Sample 103-639D-5R-3, 34 cm; polarized light; scale bar = 0.5 mm. C. Breccia of biohermal limestone. Skeletal debris is replaced by coarse sparry calcite, with silt-size peloids accumulated in interparticle voids. An unidentified encrusting alga is in the center of the photograph. Sample 103-639D-5R-3, 34 cm; polarized light with gypsum plate; scale bar = 0.5 mm. D. Un• sorted skeletal-peloid packstone with fragments of algal-coated corals, mollusks (note the development of the stromatolite in the center of the photograph), echinoids, large foraminifers, sand-size skeletal debris, and peloids. The lithified rock is fractured, with hairline frac• tures healed by sparry calcite. Sample 103-639D-5R-3, 13 cm; ordinary light; scale bar = 0.5 mm. E. Cryptalgal structure of a "deep- water" thrombolite. Spongiform microstructure resembling Spongiostromata consists of silt-size micrite clots. Filamentous texture is visi• ble at the right upper corner of the photograph. The denser laminae probable represent periods of interruption in the thrombolite growth. Similar structures are found to be associated with biohermal debris deposited on the lower reef flanks. Sample 103-639D-5R-2, 95 cm; ordinary light; scale bar = 0.5 mm. F. Silt-size peloid wackestone with debris of ostracode and calpionellid tests. Preservation of the frag• ile shells indicates a low-energy depositional environment and pelagic influence on deposition. Sample 103-639D-5R-1, 105 cm; polarized light, scale bar = 0.2 mm.

180 LATE JURASSIC CARBONATE PLATFORM zoans, echinoids, ostracodes, sponge spicules, and benthic fora• 1987b), whereas local bafflestone/framestone-like textures are remi• minifers (lituloids). Larger foraminifers (A. lusitanica; Moullade niscent of in-situ building organisms. However, this inference is et al., this volume) are also present. Scattered occurrences of not well substantiated. calpionellids are common (Fig. 7F); coarse sand to pebble size Dark, inclusion-rich dolomite is abundant in the thin sec• fragments of corals, calcisponges, and hydrozoans are rare. tions and is accompanied by some clearer, inclusion-free dolo• Trace amounts of poorly preserved debris of red algae and algal mite (largely former calcite veins). Scattered echinoderm ossi• oncoliths are also found. Allochems are deposited in a matrix of cles and ghosts of algae (dasycladaceans), foraminifers, bryozo- micrite and silt-size peloids. Geopetal fabric is well developed in ans, gastropods, hydrozoans, and corals were identified in the Sample 103-639D-4R-1, 42-44 cm. The original voids, which dolomite, but patterns reminiscent of a much richer organic could have been formed by the boring action of sponges, are component remain largely unidentified. A distinctive feature of partially filled by silt-size peloids, with the remaining part of the marbled dolomite are flat, coupled "platy shells" of a thick• the void filled by coarse sparry calcite. ness less than 1 mm, which have been previously referred to as phylloid algae (Fig. 11, "Site 639" chapter; Shipboard Scientific Unit IV Party, 1987b). However, thin-section analysis did not confirm The intervals from Sample 103-639D-4R-1, 10 cm, through this interpretation. Downhole, where the dolomitization is less Core 103-639D-2R in Hole 639D, Cores 103-639B-4R through intense (Cores 103-639B-3R-103-639B-4R and 103-639C-2R), the 103-639B-2R in Hole 639B, and Cores 103-639A-10R through dolomite is a lighter gray color and of a peloid-skeletal composi• 103-639A-8R in Hole 639A consist of dolomite. From the inter• tion, with a varied biota (Fig. 8C) and local occurrences of soli• pretation of seismic data and lithologic appearance, we divide tary coral heads. Irregular tubules penetrating Sample 103-639B- the dolomites into a brown marbled dolomite (Subunit IVB) re• 4R-1, 98 cm, may indicate that this dolomite could have been covered in Hole 639B and stratigraphically correlated for Hole part of a rhyzolith (Fig. 8D). 63 9D and a stratigraphically younger, light-colored sucrosic do• In the marbled dolomite, a multistage system of cavities and lomite (Subunit IVA) penetrated in Hole 639A (Fig. 3). fractures documents complex multiple dolomitization, as we discuss subsequently. Both blocky dolomite (former calcite ce• Subunit IVB: Brown Marbled Dolomite ment) and sediment are void-filling material. The sequence of The recovery of Subunit IVB dolomites in Hole 639D (Sam• void-filling events closely resembles that in the upper vuggy do• ple 103-639D-4R-1, 10 cm, through Core 103-639D-2R) was very lomite in Hole 639A. Cement alone is dominant in the smallest low, with recovery rates of only 2% for Core 103-639D-2R and veins. The earliest recognizable features are irregular, wide voids, 9°7o for Core 103-639D-3R. Difficulties with core recovery in reminiscent of either primary interparticle voids of a coarse Hole 639D seem to be explained by the geophysical logs, which clastic deposit or open spaces within an organic framework. A indicate that the carbonate bed present in each of the cored in• later-formed system of fractures is superimposed and seems to tervals is only 2 m thick, with the rest of the interval consisting crosscut the earlier structures. The geometric relationships be• of marls or clays. Based on this evidence and on the small frag• tween the earlier cavities and the crosscutting fractures are com• ment size of the recovered dolomite, the dolomite in Hole 639D monly obscured. Cavity formation and fracturing in the mar• could represent a dolomitized debris flow. However, because the bled dolomite predates the main dolomitization event, as docu• soft sediment intercalated with the dolomite beds was not recov• mented by the crosscutting relationships between dolomite crystal ered, we can not exclude that the dolomite is an in-situ deposit. boundaries and the fracture walls. The dolomite is yellowish brown and grayish brown, fine to coarse crystalline, and vuggy. About 7% of the irregularly shaped Subunit IVA: Light-colored Sucrosic Dolomite vugs are several millimeters in size and are scattered throughout Light brownish gray and yellowish brown, medium to coarse the rock. Some of the vugs are partially infilled by milky white crystalline, sucrosic dolomite was recovered in Hole 639A (Sec• dolospar whereas others are empty. The dolomite is cut by white tion 103-639A-8R, CC, through Core 103-639A-10R). The origi• veins, up to 5 mm thick, which are also of dolomitic composi• nal texture and composition of the limestones has been obliter• tion. The original limestone texture was completely obliterated ated by extensive dolomitization, similar to that in Subunit IVB. by wholesale dolomitization. The thin sections show medium to Some ghosts of gastropods, bivalves, and echinoderms can be coarsely crystalline, euhedral, mostly zoned dolomite crystals recognized. The other components are rounded bodies, 0.5 mm with abundant impurities (Fig. 8A). The zoned dolomite crys• in diameter, that resemble the coated grains in Core 103-639D- tals have rhombic, turbid nuclei. Iron oxides are found as par• 9R. These occur together with echinoid debris, suggesting that tial infill of some of the intercrystalline voids. The original the original carbonate was probably an echinoid-ooid(?)-skele- limestone fabric can sometimes be recognized by ghost struc• tal grainstone and packstone. tures of echinoids, mollusk debris, rare calcisponges and hydro• Open pores and sediment-filled and cemented cavities and zoans, and/or bioclasts of enigmatic origin (Figs. 8B through fractures, as large as several centimeters in diameter, are notable 8D). Extensive syntaxial cement around echinoderm fragments features in the dolomite (see fig. 9 in "Site 639" chapter; Ship• is evidence that ample interparticle space was present during ce• board Scientific Party, 1987b). Locally, cavities and fractures mentation, indicating low compaction and early lithification of are large enough to give the dolomite a brecciated appearance. the dolomite precursor. Larger open cavities, reminiscent of fossil molds, are as much as A similar dark, grayish brown, massive dolomite with minor 1-1.5 cm across. The smaller open pores are rounded in shape, intraclast breccia zones and complex patterns of voids, void fill• average 1 to 3 mm across, and occur sparsely in most of the do• ings, and fractures filled by internal sediment and cement is lomite samples (birdseye-type structures?). Whatever their size, present in Hole 639B (Cores 103-639B-2R through 103-639B- the open pores are lined by a single generation of drusy dolo• 4R). We consider it to be part of Subunit IVB. This dolomite is mite. These infilled cavities are roughly equidimensional or ir• a comparatively low-porosity rock, with only a few, large pri• regular in shape, or they display a sheetlike and/or vertical-frac• mary open pores. Typically, the dolomite shows a translucent, ture geometry. In the simplest case, void spaces are floored by a "marbled" appearance. The color shades and irregular mottling first generation of clear drusy dolomite, geopetally filled by in• are reminiscent of predolomitization structures. The rubbly ap• ternal sediment, and sutured by clear, blocky dolomite cement. pearance sometimes mimics a floatstone or rudstone precursor Cavity-lining dolomite commonly displays a baroque habit (curved (see fig. 10 in the "Site 639" chapter; Shipboard Scientific Party, crystal faces and undulose extinction; Folk and Assereto, 1974).

181 L. F. JANSA ET AL.

Figure 8. Thin-section photomicrographs of Unit IV. A. Medium crystalline dolomite with a complex dolomitization pattern. The dolo• mite precursor, probably an echinoid-oolitic packstone or wackestone, has been replaced by crystalline dolomite. Syntaxial rims of echi- noids are replaced by clear dolomite. Other finer grained carbonate sediment particles or peloids are replaced by coarse dolospar with dirty nuclei. Larger clasts or voids were replaced (filled) by a patch of multiple-zoned coarse dolospar. After the period of leaching and solution, which produced enlarged voids and channel-type porosity, these were infilled by clear dolospar. The last stage was infiltration of iron oxide-stained clay sediment deposited in intercrystalline pores. Sample 103-639D-3R-1, 32 cm; ordinary light; scale bar = 0.1 mm. B. Coarse crystalline, poorly visible zoned dolospar with enclosed enigmatic tubiform structure in the center. Sample 103-639B- 4R-1, 88 cm; polarized light; scale bar = 1 mm. C. Well-preserved ghost texture of calcisponge, micrite-coated intraclasts, and a ghost peloid texture replaced by coarse zoned dolospar in an original rock that was probably a biohermal limestone similar to that in the un• derlying core. Sample 103-639D-4R-1, 3 cm; ordinary light; scale bar = 1 mm. D. Medium crystalline dolomite, with interwoven net of tubules. Individual tubes, some branching, are bordered by a film of impurities. We interpret these as root casts. Rhyzolith is a reliable indicator of caliche and of subaerial exposure. Sample 103-639B-4R-1, 87 cm; ordinary light; scale bar = 1 mm. E. A fracture infilled by a mixture of unsorted broken fragments of dolomite crystals and dolomicrite. The wall of the fracture is rimmed by baroque dolo• mite. Note that the baroque dolomite crystals grow in an optical continuity with the dolomite crystals that were cut by the fracture. Sample 103-639B-4R, CC (12 cm); polarized light; scale bar = 1 mm. F. A fracture in the dolomite, infilled by dark dolomicrite and angular fragments of dolomite. The impure dolomite crystals and crystal zonation of the clasts in the fill is a clear documentation that this fracture was formed after the dolomitization of the limestones was completed. The fracture is rimmed by a discontinuous rim of clear dogtooth dolospar, which filled the remainder of the void after sediment compacted in the fracture. Sample 103-639A-10R-1, 81 cm; ordinary light; scale bar = 1 mm.

182 LATE JURASSIC CARBONATE PLATFORM

Baroque dolomite also is found lining fracture walls (Fig. 8E), AGE OF CARBONATES and locally, it grows diagenetically at the expense of the internal The biostratigraphic age dating of Subunit VA in Hole 639D is sediment. Internal sediment in cavities chiefly consists of light based on calpionellids (Core 103-693D-5R), of which the presence gray, intraclast-rich peloid dolomicrite, typically containing bro• of Calpionella alpina, Crassicollaria massutiniana, and Crassi- ken euhedral crystals of silt- and fine-sand-size dolomite. Crys• collaria brevis indicates a middle late Tithonian age (Crassicola- tal silt and a dolomite microspar mosaic, possibly replacing a ria A Zone) (Shipboard Scientific Party, 1987b). The deeper primary crystal silt sediment, also occur as cavity fillings. Dis• cores (103-693D-6R to 103-693D-13R) are characterized by the crete laminae of dolomitic crystal silt and fine sand are interlay- presence of A. lusitanica, indicating a Tithonian age for Subunits ered with dolomitic micrite. Internal sediment layering indicates VB and VC. This foraminifer, according to Hottinger (1967), was a multiple filling of the cavity. The presence of a cement crust found in Morocco in Tithonian and Kimmeridgian age sedi• between sediment laminae required that a certain amount of ments; however, in the Bonnition H-32 well on the Grand Banks time elapsed before the following lamina was deposited. Lami• (fig. 24 in Jansa et al., 1980; Grant et al., this volume)—juxta• nations may suggest moderate current activity within caves or a posed to Galicia Bank—A. lusitanica was not found in sedi• storm-influenced bottom. ments older than Tithonian. Thus, a Tithonian age for the sedi• A late stage of randomly oriented to subvertical cracks, a few ments enclosing A. lusitanica at Site 639 is most likely. millimeters up to 1-1.5 cm wide, crosscuts the earlier formed The dolomites in Holes 639D, 639B, and 639A generally lack structures. The fracturing is brittle and, based on the type of in• identifiable fossils for biostratigraphic dating, but they overlie filling material, we assign these features to two basic categories: limestones dated as Tithonian. Daniel and Haggerty (this vol• (1) small-scale to microscopic, hairline cracks and veins, with a ume) identify benthic foraminifers revealed by fluorescence pe• white sparry dolomite filling, and (2) larger fractures with yel• trography in Sample 103-639A-10R-2, 16-21 cm, as a Dorothia lowish dolomicrite or dolosparite fill and scattered angular chips sp. assigned to D. cornula (Reuss) and a trochamminid. These of the dolomitized host rock (Fig. 8E). In the hairline fractures, are neritic or upper bathyal forms that may be as young as Val• dolomite crystal boundaries crosscut fracture walls, showing that anginian. On the basis of stratigraphic superposition and this dolomitization postdates fracturing. The blocky habit of the biostratigraphic data, the dolomites may range in age from the dolomite suggests that a possible precursor was low-magnesium late Tithonian to Valanginian, even though interpretation of do• calcite cement. In the wider fractures, brecciation was nearly in lomitization events at Site 639 indicates that a Valanginian age situ with little displacement of the clasts (Fig. 8F). The clasts are for the top of carbonate platform is very unlikely. In view of the of similar composition and texture to the rock exposed at the age of the nearby Mesozoic carbonate platforms, we consider fracture wall. Clast boundaries and fracture walls are sharp, the dolomites to be late Tithonian to ?early Berriasian in age, in with a few showing a thin rim of clear dolospar, which suggests accordance with the age of the carbonate platforms in the Arosa that the late-stage fracturing postdates the main dolomitizing Basin and the eastern Grand Banks (Grant et al., this volume). event. Sediment was injected into the veins, intimately permeat• ing even the smallest veins; gravitational settling occurred with DEPOSITIONAL ENVIRONMENT widening of the fractures. Some of these latter fractures may represent neptunian dikes. Our interpretation of the depositional environment of the carbonates encountered at Site 639 is graphically summarized The top of the dolomite, in Section 103-639A-8R, CC, is on Figure 9. The prominent bimodality of the dominant oncol- overlain by homogeneous pale yellow marlstone. The contact ith-skeletal-peloid wackestone to packstone microfacies of Sub• between these two lithologies was not recovered, but it is proba• unit VC indicates sediment mixing from two sources. The finer bly sharp. The marlstone is described in the following in more fraction, dominated by hard, fine-sand-size peloids, is a facies detail because of its importance in the geologic history of the commonly found in the shelf lagoon environment, such as on underlying carbonate unit. the western side of Andros Island, Bahamas. The biota associ• ated with this microfacies is shallow marine and of normal sa• Unit III: Nannofossil Marlstone linity. However, the variety of the fauna in Subunit VC is less The dolomite in Hole 639A is overlain by 48.9 m of pale yel• than that in the overlying Subunit VB; in particular, small ben• low and pinkish gray, mottled marlstone, intercalated near the thic foraminifers are rare and dasycladacean algae are common. base with 10- to 20-cm-thick beds of a stiff marl, giving an ap• Such an assemblage is found in a shallow, slightly restricted ma• parent bedding inclination of 30°. Near the top of the unit, in rine environment. The fauna enclosed as oncolith nuclei are Cores 103-639A-4R through 103-639A-6R, are some sandstone large foraminifers, algae, and/or mollusk debris; the oncoliths laminae and beds up to 10 cm thick that have been highly dis• formed in a shallow subtidal to deep intertidal environment turbed by drilling and, thus, have no identifiable sedimentary (Flugel, 1982). The strong micritization of the oncoliths is envi• structures preserved. Marlstone is bioturbated and fractured, ronmentally significant; it could be the result of intensive algal with hairline fractures dipping about 60° and stained yellow by or fungal borings, but it also could be from the transformation limonite. Nannofossils and clay are the dominant components of carbonates exposed to meteoric water. Thus, a shallower de• of the marlstone, with quartz and feldspar silt grains, mica, and positional environment is most likely for the oncolithic lithofa• glauconite in trace amounts. The biogenic component, except cies at Site 639. for the nannofossils, includes Nanoconus, calpionellids (Cores The skeletal-peloid packstone to wackestone lithofacies of 103-639A-5R to 103-639A-8R), and trace amounts of foramini• Subunit VB has a richer faunal association, particularly of fora• fers, mollusk debris, and radiolarians. Dolomite rhombs occur minifers, than that of Subunit VC. Large foraminifers, proba• in traces near the base of the unit. bly reworked from a shallower environment, are found as lami• The foraminifers, calpionellids, and nannofossils indicate a nae in beds of intercalated marlstones. This may indicate that Valanginian age, with the base of this unit of middle early Val• Subunit VB was deposited in a less restricted, more seaward ma• anginian age (Moullade et al., this volume; Shipboard Scientific rine environment than Subunit VC. Such an interpretation seems Party, 1987b). The unit is a hemipelagic deposit laid down at the to be contradicted by the occurrence of reddish colored terrige• outer, deeper continental shelf or upper slope environment. The nous sandstones and marls, which would favor deposition in a sandstone intercalations near the top of the unit are probably highly oxidized ?nearshore marine environment. The presence of turbidity current deposits. fossil remains, including large foraminifers, and coated quartz

183 L. F. JANSA ET AL. Diverted river system River channel and inter-distributary bays (source of terrigenous sand and mud into carbonate sediment-dominated lagoon)

quartzose crevasse sand Peri-platform sand belt (skeletal packstone)

Inner shelf lagoon \V\0\ \ ^ rCVr-s^^v! (moderate water depth 2-5m) \0\ \0v i/A7\y}\> Oncoidal-peloidal packstone, wackestone, NsjSTNr<3-fV mudstone, /■vS^i R*IWi large forams, biohermal -' \hm A. solitary corals, mounds /VVV^ hard peloids, Gullied ramp oncoids, faults -' ^ outer ramp-slope deposits dasycladacean algae, (hemipelagic biomicrites, sporadic clay influx calcareous mudstones, from river, calpionellids) locally clayey limestone forereef coarse sketetal debris Figure 9. Block diagram portraying the depositional environmental setting of carbonates at Site 639. grains in the sandstone indicates transport of elastics into the That the echinoid-oolitic grainstone and packstone of Sub• carbonate environment. The poor sorting and mostly angular unit IVA (Hole 639A) was deposited in a higher energy, shal• shape of the clastic grains is indicative of a short transport dis• lower water environment than the limestones of Subunit IVB tance. can be interpreted either as an indication of progressive shallow• Local tectonism probably increased the gradient of some of ing of the carbonate platform surface or as the result of shelf the streams, which would have resulted in elastics being trans• edge morphology, because carbonate platforms commonly de• ported farther onto the shelf into the carbonate depositional do• velop at the shelf edge (Jansa, 1981). No indication of a hard- main. An alternative explanation that the occurrence of elastics ground was observed on the uppermost recovered dolomite core, is the response of sea-level lowering is not supported by the en• which indicates that the top of the dolomite represents an ero• closed fauna. sional surface or that the contact was not recovered. The importance of the elastics in providing relief for initia• tion of the Subunit VA bioherms cannot be established because LIMESTONE DIAGENETIC HISTORY of the limited core recovery, but a significant role appears prob• The diagenesis of the limestones is relatively simple whereas able. Furthermore, insufficient core recovery does not allow us that of the dolomites is exceptionally complex (Figs. 10 and 11) to determine whether the biohermal limestones are in-situ low- (Daniel and Haggerty, this volume). The hard peloids and on- relief mounds, a back-reef deposit, or debris from the lower coliths in the wackestone and packstones do not show any pro• flank of a larger reef structure. The presence of geopetal cavi• nounced grain-surface deformation by the overburden, which ties, microbial crusts, sponges, and intercalations of quartz silty indicates an early cementation. Similarly, geopetally filled cavi• peloid packstones and wackestones with calpionellids and marly ties require early cementation for their preservation. Extensive beds favors the interpretation of low-relief biohermal mounds, development of syntaxial overgrowths on echinoid plates fur• several tens of centimeters to several meters in height, in a mod• ther confirms early diagenetic cement precipitation. The soft erate water depth (30-50 m) near the paleoshelf edge. Similar peloids are merged but not flattened, indicating that cementa• mounds, 1 to 2 m in height, are known from exposed Upper Ju• tion occurred soon after deposition of some of the overburden. rassic limestones of the Lusitanian Basin and Algarve, studied Evidence of fibrous, fringing, early submarine, or phreatic ce• by the senior author and described in part by Ellis (1984). We ments is conspicuously lacking. Calcite cement occurs rarely have indicated previously that the dolomites in Hole 639B were and is a blocky, low-iron variety, typical of burial-type cement. originally biohermal limestones; thus, we interpret these de• Molds from dissolved bioclasts are occluded by blocky sparry posits to be part of the same stratigraphic horizon as the bioher• calcite, which is dominantly low ferroan, with some ferroan oc• mal carbonates in Hole 639D. currences. One coral near the bioherm surface was replaced

184 LATE JURASSIC CARBONATE PLATFORM with silica. The source of the silica could have been the sponges. Sheetlike and irregularly shaped cavities are interpreted to be Micrite "matrix" is extensive in these limestones and is associ• early diagenetic features resulting from the desiccation, shrink• ated with the occurrence of silt-size peloids. Considering that age, and fracturing of unconsolidated carbonate sediment that only minor compaction is observed, this part of the micrite rep• experienced repeated subaerial exposure. Leaching of the ex• resents an early diagenetic cement that probably precipitated at posed carbonate may have widened the early formed cavity sys• shallow depth (centimeters) beneath the sediment surface. tem. These structures similar to "fenestrae" (Tebbutt et al., Stylolitization resulting in chemical compaction is very in• 1965; reviewed in Fliigel, 1982) are found in frequently exposed tensive in the limestones, leading in places to the removal of sig• intertidal flat sequences, where the exposure and weathering of nificant parts of the limestone, as evidenced the different litho• sediments take place. facies occurring on opposite sides of stylolitic seams. Stylolites Birdseyelike structures preserved in the dolomites of Site 639A were also avenues for migration of solutions, as suggested by are nearly rounded, bubblelike cavities that were first described the oxidation and partial dolomitization of sediments along in shallow-water carbonate sequences, in association with de• some of the stylolites. Minor dedolomitization in Subunit VC caying algal mats (Ham, 1952; Tebbutt et al., 1965). However, has been related by Loreau and Cros (this volume) to a pre-late these structures are recognized as common to both carbonate burial diagenetic event, resulting from freshwater alteration. and siliciclastic tidal sediments (Deelman, 1972; M. Sarti, un- publ. data, 1987). Birdseyes are gas bubbles that form within the Early Diagenetic Structures in the Dolomitic Sequence sediment as a result of biotic or abiotic processes (see review in The final stages of the development of the carbonate platform Fliigel, 1982). Air bubbles developing in clastic, uncohesive, are represented by carbonate strata that were post-depositionally and organic-matter-free sediments are probably the result of air dolomitized. Some textural features were preserved, thereby pro• trapping during subaerial exposure and flooding in tidal zones viding an insight into the environment of deposition of this (Deelman, 1972). unit. Cavity patterns and their infilling provide additional clues These data indicate that the carbonate precursor to the dolo• about sedimentary and diagenetic processes. The diagenetic path• mite at Site 639 underwent subaerial exposure (shrinkage, leach• ways of the dolomite at Site 639 are diagramatically summarized ing of carbonates, and perhaps even evaporites precipitation). in Figure 10. Gas-bubble porosity identified near the top of carbonate se-

Inferred Coarse Original sediment: crystalline Calcareous bioclastic sand dolomite (+ micrite-rich sediment) —

2 Evaporite precipitation ? closed no \ Primary voids: limited _^ Open • cavity »- internal — (fossil molds, air bubbles) cementation pores system sediment

Periodical subaerial exposure: Shrinkage weathering, dissolution of evaporites

(intraclastic micrite) Early diagenetic cavity open Leached cavities, Internal sediment cement system formation: birdseyes, -cavity- internal dolomite crystal deposition precipitation microcaves, fractures system silt/fine sand

Shallow burial micrite cement precipitation

fracturing Baroque dolomite coarse (? Blocky, low Mg calcite) Pelagic marlstone blocky deposition Partial cement dolomite dolomitization

2nd late-stage fracturing dolomitization and and sediment injection fractures filling

Inferred original sediment: organism- built limestone framework, locally rudstone in micrite-rich matrix

Primary voids: interparticle pores in Open rudstone, open spaces closed within and/or between system pores frame-building organisms open internal cement system sediment precipitation

Stylolitization fracturing Baroque calcite neomorphism dolomite Coarse (? Blocky, low Mg calcite) blocky cement dolomite

Deposition Early burial diagenesis, tectonism, block faulting Platform drowning, early diagenesis ! subaerial exposure, fresh/brackish water diagenesis | late burial diagenesis

Figure 10. Diagram showing the major diagenetic pathways of dolomites at Site 639. A correlation with the structural evolution of the carbonate plat• form is shown at the base of the diagram. For discussion, see text.

185 L. F. JANSA ET AL.

quence is a good indicator of deposition in an intertidal envi• concentrated seawater in the form of oxidizing hypersaline brines ronment. The presence of vadose crystal silt in some of the cavi• originating in an evaporitic environment. ties suggests that the sediment periodically resided above the The positive carbon isotopic composition of dolomite (613C water table. The baroque dolomite observed in Sample 103- averages + 1.5) is consistent with marine sources (Haggerty and 639A-10R-2, 16-24 cm, may be a pseudomorph after evaporite Smith, this volume), but according to Loreau and Cros (this vol• minerals (Friedman, 1980), which would suggest temporary hy• ume), similar values are also found in dolomites originating by persaline conditions (see also Haggerty and Smith, this vol• freshwater/salt-water mixing. The very light oxygen isotopic com• ume). position, which exhibits a range of values from - 8 to - 11.3 for However, baroque-type dolomite also occurs as cavity-filling 5180, may point to a late diagenetic, low-temperature "metaso• cement and fracture lining (Fig. 8E), and in this case it is not an matic" origin of dolomitization (Mattes and Mountjoy, 1980). obvious replacement of evaporite minerals. Folk and Assereto Fluid-inclusion homogenization temperatures ranging from 67° to (1974) typically refer this type of dolomite to dolomitization in 84°C (Haggerty and Smith, this volume) support such an inter• the mixing zone of ground and saline waters. Its isotopically pretation. The latter authors interpret the oxygen isotope data light 5180 signature (Haggerty and Smith, this volume) indicates and homogenization temperatures to be the result of shallow- that this dolomite could form from hypersaline brines associ• burial dolomitization by oxidizing, hypersaline brines associ• ated with an evaporative loop during shallow burial or by re• ated with a higher geothermal gradient. Alternatively, they equilibration during deep burial, although Loreau and Cros could be a consequence of a diagenetic overprint during deeper (this volume) use the same values to argue for a freshwater in• burial with concurrent reequilibration of the oxygen isotopic fluence. composition of the dolomite, as well as the homogenization Broken, euhedral crystals of clear dolomite and crystal silt temperatures of the fluid inclusions (Haggerty and Smith, this are common components of the internal sediment fill of frac• volume). The homogenization temperatures, without pressure tures and microcaves at Hole 639A (Fig. 8F). These broken crys• correction and assuming the inclusions have not stretched dur• tals represent reworked fragments of dolomite crystals from a ing deeper burial, can be used as a very crude dipstick to test the poorly lithified sucrosic dolomite at the top of the platform. conditions at which the dolomite crystals were formed. If we They are interlayered with laminae of clayey dolomicrite. Some consider the temperature of the evaporitic environment at the of the voids are filled with yellowish clay and silt. The isotopic top of carbonate platform to have been 35°C (McKenzie et al., composition of the yellow silt, with 513C ranging from + 2 to 1980), then a rise of 20°-30°C would have resulted from burial. + 3 and 5180 from - 4 to +3 (Loreau and Cros, this volume), is Assuming a geothermal gradient of 35°C/1000 m, it can be ar• indicative of early diagenetic dolomite replacing marine carbon• gued that the main dolomitization event may have occurred in a ate. This dolomite represents the second(?) stage of dolomitiza• limestone sequence upon burial by approximately 500 m of sedi• tion that occurred immediately after initial drowning of the car• ments. The subsequent occurrence of several minor episodes of bonate platform. dolomitization is associated with fracture filling and the filling of secondary, leached voids. Late Fracturing Dolomitization Although the yellowish sediment that fills the late-stage frac• Interpretation of the petrographic, tectonic, geochemical, and tures is itself dolomitic and dolospar rims are developed around isotopic data for the dolomites of Site 639 results in contradic• some of the clasts filling fractures (Fig. 8F), we concluded that tory interpretations of the dolomitizing event (compare with Lo• no major whole-rock recrystallization took place after the latest reau and Cros, this volume; Haggerty and Smith, this volume). set of fractures was formed. Mild replacement occurred as a re• The Valanginian marlstone separated by an unconformity from sult of contact with the dolomitic host rock, in a similar fashion underlying carbonate shows minor signs of dolomitization. Iso• to what is envisaged for the overlying pelagic drape. Some of the lated dolomite crystals were observed to have grown diageneti- yellowish sediment is gravitationally settled within the cracks, but cally into the sediment near the base of the formation, but no other occurrences seem to be injected as a result of fracturing, massive replacement has occurred. This late mild metasomatism implying that the late-stage fracturing postdates both the major was affected by contact with the dolomite and circulating for• dolomitization event and the deposition of post-carbonate-plat• mation water, postdating the main dolomitizing event (see "Site form sediments. 639" Chapter; Shipboard Scientific Party, 1987b). The brittle This evidence from the rock record demonstrates that a long- fractures healed by calcite and then dolomitized together with lasting episode of brittle fracturing is sandwiched between pos- the host rock indicate that early fracturing and rock lithification tlithification and immediate predolomitization and post-early predate dolomitization and consequently, predate deposition of Valanginian platform drowning, encompassing a period of about the Valanginian marls. Dolomitization is postfracturing, which 4 m.y. Different types of fracture filling formed depending upon suggests that it occurred on an already fault-dissected margin. the diagenetic environment in which the fracturing occurred. Less intensive dolomitization in Hole 639D is best explained by Sediment-free brittle fractures formed early, when the pelagic the different elevation of the site in respect to the dolomitization cap was absent. These fractures are most probably Tithonian in front. The downward-extending dolomitization front and the age. Sediment-injected fractures formed later, after the pelagic distribution of the dolomite across stratigraphic boundaries im• sediment had been deposited, and we conclude that this forma• ply that one viable mechanism of dolomite formation could be tion was probably during the Valanginian. the mixing of dolomitization fluids (Badiozamani, 1973; Folk We relate fracturing to tectonic processes at the margin. The and Land, 1975). If this is the case, the dolomitization of car• first period of faulting (Fig. 11) resulted in downfaulting of the bonates at Site 639 must have happened early, soon after the es• carbonate platform edge, as indicated by the varying lithofacies tablishment of a regional fluid-mixing stage, and would have in the studied holes of Leg 103. The second period of faulting is been completed before the drowning of the platform. The geo• poorly constrained, but it is indicated by exposure and possible chemical data (Haggerty and Smith, this volume) do not sup• erosion of the carbonate platform during Berriasian time. Fault• port freshwater mixing. Trace amounts of sulfate mineral inclu• ing and fracturing of the carbonate platform is seen as provid• sions in the dolomite suggest that dolomitization occurred by ing conduits through which hypersaline brines migrated and do-

186 LATE JURASSIC CARBONATE PLATFORM

Site 639 STAGE 1: TITHONIAN Landward Shallow-water deposition; sandy sea level and muddy facies, subaerial and Echinodermal (skeletal) calcarenite f^T^J^ vadose diagenesis, and local Biohermal limestone evaporite nodules

Site 639 STAGE 2: BERRIASIAN Continued subsidence and sea level f Continuing deposition, shallow-water deposition, overburden partially removed shallow burial cementation, between stages 2 and 4. neomorphism, pre-dolimitization fracturing and cementation, %:'^^ and minor faulting

STAGE 3: SHALLOW BURIAL DOLOMITIZATION Platform breakup, faulting, subaerial exposure of carbonate at uptilted egde, fluids mixing or brines replacing and dolomitizing, erosion, ? karst topography; dolomite silt infilling cavities. Mixing zone or brine seepage

STAGE 4: EARLY VALANGINIAN Site 639 Platform collapse and drowning, sea level backfaulting and tilting, subsidence, post-dolomitization fracturing and sediment injection, and minor dolomitization of pelagic drape. ^J^^^Qj^/^9^^ ^~^~ Pela9'c sediment draPe

Upper vuggy dolomite Lower "marbled" dolomite Limestone Figure 11. Basic steps of the structural evolution of the carbonate platform at Site 639, as inferred from the diagenetic history of the dolomites. For discussion, see text. lomitized the limestone (see discussion by Haggerty and Smith, We have indicated that the carbonates encountered at Site this volume). The third set of fractures, recognized by sediment 639 (Fig. 13) represent a carbonate platform. This interpretation injection in the previous fractures, indicates that movement along is supported by the thickening of the carbonates in an offshore faults persisted long after the drowning and burial of the car• direction to Hole 639A (Fig. 2). Less conclusive evidence is that bonate platform marginal blocks. Alternatively, some of these some of the carbonates in Hole 639D may represent slope debris later faults could be the result of another major tectonic period and that the top of the platform was a shoal. According to inter• during which the whole tectonic blocks were downfaulted and pretation of the seismic data (Fig. 2), the carbonates uncon• rotated eastward in the ?late Valanginian. formably overlie clastic rocks. The maximum thickness of the GEOLOGIC HISTORY OF THE CARBONATE carbonate platform could be 420 m, based on a compressional velocity of 4.5 km/s. Because clastic beds are intercalated with PLATFORM the carbonates, the thickness of the platform is accordingly less. The geologic history of the carbonate platform is recon• According to our estimates (Fig. 3) 287 m of the carbonate se• structed from the data presented in the preceding and from aux• quence was penetrated at Site 639. The carbonates are of Titho• iliary evidence from synchronous carbonate platforms formed nian to ?early Berriasian age. From the seismic data, it is not during the Late Jurassic on the northern central North Atlantic improbable that the carbonates may extend down into Kim• margins. meridgian.

187 L. F. JANSA ET AL.

The depositional setting for the carbonates was a shelf lagoon nian carbonate platform of the Galicia Bank is located between (Fig. 9) comparable to the western side of Andros Island, Baha• the platforms of the Arosa Basin and of the Carson Sub-basin. mas, or an inner shelf environment similar to that of northeast This suggests that the Galicia carbonate platform was probably Australia. Clastics intercalated within the limestone sequence, part of a broader carbonate shelf that evolved during the Titho• poor sorting and angular grain shape, and a high concentration nian (Fig. 12). The intercalations of limestones and shales in the of feldspars point to a low-relief, continental, acidic-type (gran- Upper Jurassic of the Carson Sub-basin are similar to those in odiorite-dominated) basement, subaerially exposed not far from Hole 639D (Subunit VC), denoting in-situ carbonate deposition the depositional site. Thus, the carbonate platform at the west• for the hole. Furthermore, the similarity in lithologic composi• ern side of Galicia Bank could have been a localized rimmed tion of the Tithonian limestones between Galicia Bank, Carson platform, developed around a low-relief island formed atop the Sub-basin, and Meriadzek Terrace suggests that the sea covering basement block. the carbonate shelf was very shallow. The recovered cores do not provide sufficient information A disruption in deposition occurred in all three areas very about more subtle environmental changes in the platform devel• early in the Cretaceous, with the shortest disruption on the opment. The facies indicate progressive deepening as result of Grand Banks and longest one at the Meriadzek Terrace. Only Tithonian transgression in Hole 63 9D, which reversed during on Galicia Bank has intensive dolomitization been observed. the late Tithonian to a shallowing trend, in turn leading to the What makes the western margin of the Galicia Bank different development of shallow, high-energy "skeletal-oolitic shoals" from other carbonate platforms is the extensive fault tectonics, that buried the bioherms (Hole 639A). This change correlates development of a postdepositional unconformity, and, with less with the eustatic sea-level lowering that occurred just prior to certain identity, subaerial exposure and erosion. Geochemical deposition of the Crassicolaria A Zone, according to Haq et al. and isotopic data clearly indicate the importance of postdeposi• (1987). Further support for eustatically controlled environmen• tional fluids on dolomitization at Site 639. The observed 5lsO- tal change can be found in the history of Late Jurassic carbon• depletion shift from meteoric water and petrographic data (Fig. ate platforms at the circum-North Atlantic margins. 13) discussed previously point to the dolomitization as being During the Late Jurassic, the development of carbonate plat• late diagenetic and low-temperature and that it occurred soon forms along the North Atlantic margins become extensive (Jansa, after tectonic faulting began, when the platform was still buried 1981; Jansa and Wiedmann, 1982). Because such development under several hundred of meters of sediment. The fluid path• was synchronous on the separated continental plates (African, ways allowed oxidizing, sulfate-bearing, evaporitic brines to dis• North American, and European), climate and eustatics were place less saline and less dense pore fluids during shallow burial probably the principal controlling factors in the evolution of the of the platform in a higher geothermal-gradient regime associ• platforms. The most proximal carbonate platform eastward of ated with rifting, which aided the dolomitization process. The Galicia Bank is on the eastern flank of the Arosa Basin. The se• source of the evaporitic brines was the localized evaporitic envi• quence is about 400 m thick and is dominated by shallow-water ronment, which developed as the result of a combination of tec• carbonates. The similar thicknesses of these two platforms, even tonics (faulting) and eustatic sea-level lowering. An undeter• with the Galicia Bank platform located 200 km west of the mined thickness of sediments was removed at the unconformity Arosa Basin, supports our interpretation that the carbonate during subaerial exposure prior to the platform drowning. This mass at Site 639 is part of a platform and not a ramp, as sug• erosion could have been accentuated by a major eustatic sea- gested by Moullade et al. (this volume). To the north, at Deep level drop that occurred during the early Valanginian (126 Ma; Sea Drilling Project Site 401 on the Meriadzek Terrace in the Haq et al., 1987). After that the platform was drowned, dated Bay of Biscay, 94 m of Upper Jurassic to Lower Cretaceous pel- paleontological^ as prior to deposition of Calpionellid Zone E, let-intraclast grainstone and packstone, containing some debris the middle part of the lower Valanginian (Moullade et al., this of corals, ?hydrozoans, and rare calpionellids, was recovered volume). It remains unclear as to whether this initial carbonate (Monadert, Roberts, et al., 1979); this lithology is similar to platform drowning is the result of eustatic sea-level rise or from that encountered at Site 639. Limestones at this locality are un• a renewed period of tectonism as the rifting of the outer Galicia conformably overlain by upper Aptian chalk. The third area of margin progressed. concern is the Carson Sub-basin on Grand Banks (Jansa and One of the main results of our diagenetic and dolomitization Wade, 1975; Jansa et al., 1980; Grant et al., this volume) where studies is that they provide a time constraint on the dating the Kimmeridgian-Tithonian neritic limestones and shales were pene• major tensional period on western margin of the Galicia Bank. trated during oil exploration. The Tithonian section encoun• The difference in intensity of dolomitization between Holes tered in the Bonnition well is 200 m thick and consists of shal• 639A and 639D indicates that only minor faulting occurred low-water limestone overlying medium gray, calcareous shale at prior to the main post-late Tithonian dolomitization. The ma• a sharp basal contact (see Fig. 13 in Grant et al., this volume). jor tectonic episodes are concurrent and in part postdolomitiza- Upward in the section, the limestone becomes intercalated with tion. The first tectonic episode, which is represented by block marls containing large foraminifers {A. lusitanica) and calpio• faulting, subaerial exposure, and partial erosion of the carbon• nellids. Another bed of skeletal packstone with coral debris and ate platform, is not younger than early Valanginian and not minor oolitic grains and oncoliths occurs at the top of the older than late Tithonian. Considering that an undetermined Tithonian. This upper limestone is overlain at a sharp contact thickness of sediments was deposited on the top of the carbon• by brownish gray calcareous shale, with minor silty mudstone ate platform and later removed at the erosional unconformity, beds. The shale, which contains scattered calpionellids, is of the time of tectonism can be further bracketed to ?late Berria- Berriasian-Valanginian age (Jansa et al., 1980). Even though sian-earliest Valanginian. The second period of tectonism is there is a sharp lithologic boundary at the top of the limestone, poorly established by drilling results, but can be better identi• indicating perhaps a sharp environmental change and possibly a fied by seismic data (Fig. 1). The tectonism that resulted in diastem, paleontology suggests continuous deposition. How• downfaulting and rotation of the tectonic blocks to the east oc• ever, paleontological evidence for the presence of Berriasian at curred after deposition of seismic Unit 5, which was interpreted this locality is weak. by the Shipboard Scientific Party (1987b) as corresponding to Several conclusions can be derived from the regional distri• upper Valanginian turbidites, and prior to or during deposition bution of carbonate platforms of the Late Jurassic. The Titho• of untested seismic Subunit 4 (Fig. 1 and the wedge-shaped unit

188 LATE JURASSIC CARBONATE PLATFORM

p Flemish Cap

■Galicia Bank

-U Spanish Meseta W*^ -./-/. 200 i km

^^s? Carbonate shelf and bioherms '/'. Carbonate platform

Basinal marls and pelagic limestones

Paleozoic basement ggg Shelf edge ^fSl^l si°pe

Basin

~0S Deep-sea fan Figure 12. Paleogeographic reconstruction of the Iberia-Grand Banks region during the Tithonian.

underlying reflector V on Fig. 2). The age of this unit from the Richerche (Grants No. 86.00995.05 and 87.00707.05). L. Jansa acknowl• seismic and drilling data is reasonably constrained as Hauteriv• edges the support of the Geological Survey of Canada in this study and ian. is thankful to G. Grant for drafting and N. Koziel for typing of the Miocene sediments overlie the downfaulted limestone blocks, manuscript. J. Haggerty thanks JOI-USSAC for the financial support of her research. This is Geological Survey of Canada Contribution no. except at Hole 639C, where the dolomites are overlain by red• 2004. dish brown unfossiliferous clay. A similar clay is widespread in the central North Atlantic, where it was described as Plantage- REFERENCES net (Jansa et al., 1979) of late Cenomanian to Eocene age. Such Badiozamani, K., 1973. The Dorag dolomitization model; application shale is deposited in deep water, below the carbonate compensa• to the Middle Ordovician of Wisconsin. J. Sediment. Petrol., 43: tion depth. The occurrence of such a shale at Hole 639C indi• 965-984. cates that the downfaulted edges of the carbonate platform sub• Boillot, G., Winterer, E. L., et al., 1987. Proc. ODP, Init. Repts., 103: sided to abyssal depth during the Late Cretaceous to early Ter• College Station, TX (Ocean Drilling Program). tiary. Deelman, J. C, 1972. On mechanism causing birdseye structures. Neues Jahrb. Geol. Palaeontol. Abh., 12:582-595. ACKNOWLEDGMENTS Dupeuble, P.-A., Boillot, G. and Mougenot, D., 1987. Upper Jurassic- The authors express their gratitude to A. Edwards for assistance in lowest Cretaceous limestones dredged from the western Galicia mar• seismic data interpretation and to two unknown reviewers for helpful gin. In Boillot, G., Winterer, E. L., et al., Proc. ODP, Init. Repts., critical comments that improved the manuscript. We are indebted to the 103: College Station, TX (Ocean Drilling Program), 99-106. ODP shipboard staff for their assistance during Leg 103 and to G. Boil• Ellis, P. M., 1984. Upper Jurassic carbonates from the Lusitanian Ba• lot and E. L. Winterer for encouragement and advice through the cruise. sin, Portugal, and their subsurface counterparts on the Nova Scotia Financial support for M. C. Comas was provided by the Spanish Com• Shelf [Ph.D. thesis]. Open Univ., Milton Keynes, U.K. mittee for the ODP and Consejo Superior de Investigaciones Cientifi- Fliigel, E., 1982. Microfacies Analysis of Limestone: Berlin (Springer- cas. M. Sarti was financially supported by the Consiglio Nazionale delle Verlag).

189 L. F. JANSA ET AL.

Folk, R. L., and Assereto, R., 1974. Giant aragonite rays and baroque margins: a comparison. In von Rad, U, Hinz, K., Sarnthein, M., white dolomite in teepee fillings, Triassic of Lombardy, Italy. Annu. and Seibold, E. (Eds.), Geology of the Northwest African Continen• Meet. Program Soc. Econ. Paleont. Mineral., 34-35. tal Margin: Berlin (Springer-Verlag), 215-269. Folk, R. L., and Land, L. S., 1975. Mg/Ca radio and salinity: two con• Mattes, B. W, and Mountjoy, E. W, 1980. Burial dolomitization of the trols over crystallization of dolomite. AAPG Bull., 59:60-68. Upper Devonian Miette Buildup, Jasper National Park, Alberta. In Friedman, G. M., 1980. Dolomite is an evaporite mineral: evidence Zenger, D. H., Dunham, J. B., and Ethington, R. L. (Eds.), Con• from the rock recorded and from sea-marginal ponds of the Red cepts and Models of Dolomitization: Spec. Publ. Soc. Econ. Pale• Sea. In Zenger, D. J., Dunham, J. B., and Ethington, R. L. (Eds.), ontol. Mineral., 28:259-297. Concepts and Models of Dolomitization: Spec. Publ. Soc. Econ. McKenzie, J. A., Hsu, K. J., and Schneider, J. F, 1980. Movement of Paleontol. Mineral., 28:200-203. subsurface water under the sabkha, Adu Dhabi, U.A.E., and its re• Ham, W. E., 1952. Algal origin of the "birdseye" limestone in the Mc- lation to evaporite-dolomite genesis. In Zenger, D. J., Dunham, Lish Formation. Proc. Okla. Acad. Sci., 33:200-203. J. B., and Ethington, R. L. (Eds.), Concepts and Models of Dolo• Haq, B. J., Hardenbol, J., and Vail, P. R., 1987. Chronology of fluctu• mitization: Spec. Publ. Soc. Econ. Paleontol. Mineral., 28:11-30. ating sea levels since the Triassic. Science, 235:1156-1167. Montadert, L., and Roberts, D. G., 1979. Init. Repts. DSDP, 48: Wash• Hottinger, L., 1967. Foraminiferes imperfores du Mesozoi'que Moro- ington (U.S. Govt. Printing Office). cain. Notes Mem. Serv. Geol. Morocco, 209. Montadert, L., Winnock, E., Delteil, J. R., and Grau, G., 1974. Conti• Jansa, L. F, 1981. Mesozoic carbonate platforms and banks of the nental margin of Galicia-Portugal and Bay of Biscay. In Burk, A., eastern North American margin. Mar. Geol., 44:97-177. and Drake, CL. (Eds.), Geology of Continental Margins: Berlin Jansa, L. F, Enos, P., Tucholke, B. E., Gradstein, F M., and Sheridan, (Springer-Verlag), 323-342. R. E., 1979. Mesozoic-cenozoic sedimentary formations of the Mougenot, D., Capdevila, R., Palain, C, Dupeuble, P.-A., and Mauf• North American Basin; western North Atlantic. In Talwani, M., fret, A., 1985. Nouvelles donnes sur les sediments ante-rift et le so• Hay, W., and Ryan, W.B.F. (Eds.), Deep Drilling Results in the At• cle de la marge continentale de Galicia. C. R. Acad. Sci. Ser. 2, 301: lantic Ocean: Continental Margins and Paleoenvironment: Am. 323-328. Geophys. Union, Maurice Ewing Ser., 3:1-57. Shipboard Scientific Party, 1987a. Introduction, objectives and princi• Jansa, L. F, Remane, J., and Ascoli, P., 1980. Calpionellid and forami- pal results: Ocean Drilling Program Leg 103, west Galicia margin. niferal-ostracod biostratigraphy at the Jurassic-Cretaceous bound• In Boillot, G., Winterer, E. L., et al., Proc. ODP, Init. Repts., 103: ary offshore eastern Canada. Riv. Itai. Paleont., 86:67-126. College Station, TX (Ocean Drilling Program), 3-17. Jansa, L. F, Steiger, T. H., and Bradshaw, M., 1984. Mesozoic carbon• , 1987b. Site 639. In Boillot, G., Winterer, E. L., et al., Proc. ate deposition on the outer continental margin off Morocco. In ODP, Init. Repts., 103: College Station, TX (Ocean Drilling Pro• Hinz, K., Winterer, E. L., et al., Init. Repts. DSDP, 79: Washington gram), 409-532. (U.S. Govt. Printing Office), 857-891. Tebbutt, G. E., Conley, C. D., and Boyd, D. W, 1965. Lithogenesis of a Jansa, L. F, and Wade, J. A., 1975. Geology of the continental margin distinctive carbonate rock fabric. Contrib. Geol., 4:1-13. off Nova Scotia and Newfoundland. In van der Linden, W.J.M., and Wade, J. A. (Eds.), Offshore Geology of Eastern Canada, 2: Re• gional Geology: Geol. Surv. Can., 74-30:51-106. Date of initial receipt: 23 March 1987 Jansa, L. F, and Wiedmann, J., 1982. Mesozoic-cenozoic development Date of acceptance: 21 December 1987 of the eastern North American and northwest African continental Ms 103B-122

190 LATE JURASSIC CARBONATE PLATFORM

639 B 639 A UNDETERMINED UNDETERMINED L.VALANG. AGE FOSSI L [C23 | FORAMINIFERA AGE S CALP. E. Z CALPIONELLIDS CALC. O. Z NANNOFOSSILS PALYNOMORPHS & DINOFLAGELLATES UNIT IV UNIT III LITHOSTRATIGRAPHIC UNITS 4872.3 4823.5 4814.0 4811.0 4804.6 4794.9 4785.3 4775.7 DEPTH (metres) 2C 4B 3B_ 2B 10A^ 9A_ 8A__ 7A__ CORE NUMBER & RECOVERY n 1 n TI TI Mill THIN SECTION E LITHOLOGICAL COLUMN II I I1 il Light Gray ■ Gray Dark Gray P Yellow Brown s TERRIGENOU S 1 1 1 I ONSTITUENT S ■ QUARTZ % Authigenic (A) + FELDSPAR % + CLAY / MICA (M)%

• 5 «m - 60 um GRAI N SIZ E • 60 um - 500 wm 500 um - 2000 urn > 2 mm

HX + +'x -r OPAQUES, PYRITE (P), Fe2Ox (F) CARBONATE LITHOCLAST, DOLOMITE (D) FRAME ­ TEXTURE S + ++- ++++ x BIOCLAST WOR K COMPON ­ x INTRACLAST ENT S x •► ► PELOIDS + •< ONCOIDS ► ► MICRITE/MICROSPAR (M) • • CLAY/SILT (S)

n DEPOSIT ­ GRAINSTONE PACKSTONE IONA L WACKESTONE MUDSTONE a BOUNDSTONE

SPAR RIM(S) , BLADED (B) CEMEN T + + SYNTAXIALRIM n BLOCKY FABRIC S SPARITE DIAGENETI C MICROSPAR m m con m -m n mmr n mmrr E,ln,B(1) DOLOMIT E J > 60% • — 35 - 60% 1 10 - 35% X X < 10% ► 03 « ► •< POROSITY (2) • STYLOLITES BIOTURBATION o a a CD FRACTURE FILLING (3) i GEOPETAL INFILLING (4) MICRITIZED GRAINS LITUOLIDAE BENTHI C FORAMINIFER A ANCHISPIROCYCLINA LUSITAN. LENTICULINA EPISTOMINA PSEUDOCYCLAMINA LITUS TROCHOLINA + NAUTILOCULINA OOLITICA EVERTICYCLAMMINA CONICOSPIRILLINA BASILIENSIS MILIOLIDAE

BISERIATED BENTHI C FAUN A P + + ^ + + ■H FORAMINIFERA (Unspecified) (5) x o DASYCLADACEA(6) SALPINGOPORELLA ANULATA ALGA E THAUMATOPORELLA PERMOCALCULUS ZERGATELLA + ENCRUSTED ALGAE (CYANOPH.) + + ALGAE (Unspecified) (7) + SPONGE, SPICULES (S) I III STROMATOPORIDAE - HYDROZOA ■ 1 CORALS 1 BRYOZOA + 2 GASTROPODA (NERINEA (N) (8) PELECYPODA CRINOIDEA i in i + -H + ECHINOIDEA ! + OSTRACODA SERPULIDS CRUSTACEA, COPROLITES ELAGI C ++ CALPIONELLIDS FAUN A RADIOLARIA

X> X + NANNOFOSSILS PELECYPODA CONTINENTAL BRACKISH DEPOSITIONAL NERITIC ENVIRONMENT \ EPIBATHYAL * Figure 13. Geologic chart for the carbonate sequence at Site 639. Plot shows lithology, unit bounda• ries, ages, and composition (components and textural parameters) derived from thin section studies.

191 L. F. JANSA ET AL.

TITHONIAN LATE TITHONIAN UNDETERMINED

ANCHISPOROCYCLINA LUSITANICA CRASSICOLLARIA A ZONE

SUBUNIT 5C SUBUNIT SB | SUBUNIT 5A | 5025.9 5016.3 5006.6 4997.0 4987.4 4977.7 4968.1 4958.5 4949.0 4939.3 4929.8 4922.1 13D 12D__ 10D_ 9D SD 6D 4D 2D 11Dimrr 7Dm 5DTTTT1 3D i Hill I in iTTT ilHT f » | Z3 i i 1 [1 1 1 1 I 1 1 1 1 1 znii 1 I II Z3 1 I I' 1 1 ■ 1 "1 4 + 1i + 1++ in i i 11 lllg; III. 1 i i II II i i i 1 II n 11 I"" 1 "" I ■" i- -n l-n 1- 1 1 III ■n 1 hi X XX •► ► • ► ■ ►*■ ►

• ••x• • • i x • + XX > » «

• • ► • • ■• ► • X • ► ► ► »>► ►>. ..►• ii 1 III •I i• .III ».»II 1 I•I 1 « 1 ..!.. 1 2 ►s- III ii : "IS­ ► .►► in • 1 1 II HII II 1 1 1 1 1 1 1 IIII 1 1 II 11 11 1 I in 1 1 II 11 1 ' 1 I I

►

• X x

.1 1 •1 x ► • 1 • • ► « 1. •< • I m m-~ 1-3 ,m ■

► 1 II 1 1 II ' II m 1 1 1 II in III II II • 1 11 II 1 1 I II i ' M" 1 1 ' lo O a

1 1 1 1IIII in 1 III I II i II 11 1 III i II II i i 1IIII H i 1 II 1 II un i II 1II 1 llll 1 1 1 1 + II n 1 1 1 1 1 11 I in 1 LEGEND 1 1II 1 11 1 II 11 1 1 1 i 1 II 1 M < 1% (or Trace) + 1 1 - 5% x i 1 11 II1 5 - 20% 1II 1 i 1 1 MM 11 20 - 50% « 50 - 100% m ^m31 3J i 3} 33 30 x • + PRESENT _ III 1 II II 1 II 'o' '6 o on 'o' o i 1 1 1 1 1 1

(1) DOLOMITE: E Equant anhedral In Inclusion rich B Blocky

n n Q 0 lo 0D |o aa O000 o o a (2) POROSITY: o°o °o M 0) 1 1 » IM co c/> S Secondary II 1 1 1 III .1 II (3) FRACTURE FILLING: (4) GEOPETAL INFILLING: 1 II II III 1 1 Bq Baroque 1 1 1II II 1; z iz i un Z iz ' z ZS £ 2 D Dolomite l 1 1 11 n 1IIII HI 1 llll i il 1 I Internal sediment 11 i 1 llll li II1 in 1 C Calcite 1II1 i i 1 llll i i II II II II II 1 (5) FORAMINIFERA: 11 n 1 un i R Rectocyclammina 1 i HI 1 ' E Encrusting foram. 1 1 (6) DASYCLADACEA: 11 1 1 II II 1 C Codiaceae 1 II i III 1 II (7) ALGAE (Unspecified): 1 G Gymnocodiaceae '•Li 1 C Cyanophycae S Solenoporacea

(8) GASTROPODA: M Mollusca

Figure 13 (continued).

192